A cascade eye diseases screening system with interpretability and expandability in ultra-wide field fundus images: A multicentre diagnostic accuracy study.

Background: Clinical application of artificial intelligence is limited due to the lack of interpretability and expandability in complex clinical settings. We aimed to develop an eye diseases screening system with improved interpretability and expandability based on a lesion-level dissection and tested the clinical expandability and auxiliary ability of the system. Methods: The four-hierarchical interpretable eye diseases screening system (IEDSS) based on a novel structural pattern named lesion atlas was developed to identify 30 eye diseases and conditions using a total of 32,026 ultra-wide field images collected from the Second Affiliated Hospital of Zhejiang University, School of Medicine (SAHZU), the First Affiliated Hospital of University of Science and Technology of China (FAHUSTC), and the Affiliated People's Hospital of Ningbo University (APHNU) in China between November 1, 2016 to February 28, 2022. The performance of IEDSS was compared with ophthalmologists and classic models trained with image-level labels. We further evaluated IEDSS in two external datasets, and tested it in a real-world scenario and an extended dataset with new phenotypes beyond the training categories. The accuracy (ACC), F1 score and confusion matrix were calculated to assess the performance of IEDSS. Findings: IEDSS reached average ACCs (aACC) of 0·9781 (95%CI 0·9739-0·9824), 0·9660 (95%CI 0·9591-0·9730) and 0·9709 (95%CI 0·9655-0·9763), frequency-weighted average F1 scores of 0·9042 (95%CI 0·8957-0·9127), 0·8837 (95%CI 0·8714-0·8960) and 0·8874 (95%CI 0·8772-0·8972) in datasets of SAHZU, APHNU and FAHUSTC, respectively. IEDSS reached a higher aACC (0·9781, 95%CI 0·9739-0·9824) compared with a multi-class image-level model (0·9398, 95%CI 0·9329-0·9467), a classic multi-label image-level model (0·9278, 95%CI 0·9189-0·9366), a novel multi-label image-level model (0·9241, 95%CI 0·9151-0·9331) and a lesion-level model without Adaboost (0·9381, 95%CI 0·9299-0·9463). In the real-world scenario, the aACC of IEDSS (0·9872, 95%CI 0·9828-0·9915) was higher than that of the senior ophthalmologist (SO) (0·9413, 95%CI 0·9321-0·9504, p = 0·000) and the junior ophthalmologist (JO) (0·8846, 95%CI 0·8722-0·8971, p = 0·000). IEDSS remained strong performance (ACC = 0·8560, 95%CI 0·8252-0·8868) compared with JO (ACC = 0·784, 95%CI 0·7479-0·8201, p= 0·003) and SO (ACC = 0·8500, 95%CI 0·8187-0·8813, p = 0·789) in the extended dataset. Interpretation: IEDSS showed excellent and stable performance in identifying common eye conditions and conditions beyond the training categories. The transparency and expandability of IEDSS could tremendously increase the clinical application range and the practical clinical value of it. It would enhance the efficiency and reliability of clinical practice, especially in remote areas with a lack of experienced specialists. Funding: National Natural Science Foundation Regional Innovation and Development Joint Fund (U20A20386), Key research and development program of Zhejiang Province (2019C03020), Clinical Medical Research Centre for Eye Diseases of Zhejiang Province (2021E50007).

SOX2 maintains the stemness of retinoblastoma stem-like cells through Hippo/YAP signaling pathway.

PURPOSE: To explore the mechanisms underlying stemness maintenance of retinoblastoma (RB) stem cells (RSCs). METHODS: The retinoblastoma stem-like cells (RSLCs) were isolated by single cell cloning in combination of examination of sphere-forming capacities. The stemness of the cells were characterized by the sphere-forming capacity and the expression levels of RSCs markers. Gene manipulation was performed by lentivirus system. Transcriptional regulation was identified by qRT-PCR, luciferase reporter, nuclear run-on and DNA pull-down assay. Spearman analysis was employed for correlation analysis of genes in tumor tissues of RB patients. RESULTS: The isolated RSLCs exhibited enhanced sphere-forming capacity and constantly higher levels of CD44, ABCG2, SOX2 and PAX6, but not CD133. SOX2 positively regulated the stemness of RSLCs. SOX2 directly binds to the promoters of WWTR1 and YAP and transcriptionally activates WWTR1 and YAP. Knockdown of WWTR1 or YAP partially abolished the effect of SOX2 on the stemness of RSLCs. CONCLUSIONS: SOX2, as a key deriver, maintains RB stemness by activating Hippo/YAP signaling. Inhibition of Hippo/YAP signaling would be an effective strategy for human RB caused by SOX2 upregulation.

Protective Role of Hinokitiol Against H2O2-Induced Injury in Human Corneal Epithelium.

PURPOSE: We recently found that hinokitiol has anti-inflammatory activity in human corneal epithelial (HCE) cells. Herein, we investigated the protective role of hinokitiol against H2O2-induced injury in HCE cells and the mechanisms that underlie its action. METHODS: HCE cells were incubated with different concentrations of hinokitiol or dimethylsulfoxide (DMSO), which served as a vehicle control, before H2O2 stimulus. The cell viability was evaluated using a cell counting kit-8 (CCK-8) assay. TUNEL, phosphorylated histone γH2A.X, cleaved caspase-3 expression analyses, and location of cytochrome c were conducted to detect cell injury and apoptosis. Reactive oxygen species (ROS), catalase (CAT), superoxide dismutase (SOD), methane dicarboxylic aldehyde (MDA), and total antioxidative capacity (T-AOC) were used to determine oxidative stress. Bcl-2 and Bax protein expressions were measured by western blotting. RESULTS: Hinokitiol significantly improved the cell viability, decreased the apoptosis rate, inhibited DNA damage, and reduced cleaved caspase-3 expression and the leakage of cytochrome c from mimitochondrion to cytoplasm of HCE cells against the oxidative stress induced by H2O2. Generation of ROS and MDA and decreased activity of CAT, SOD, and T-AOC were also ameliorated by hinokitiol administration. Moreover, Bcl-2 expression was down-regulated while Bax was up-regulated by H2O2 stimulus, which were reversed by hinokitiol application. CONCLUSION: Hinokitiol protects HCE cells against H2O2-induced injury likely by its antioxidant activity and modulating the Bcl-2 signaling pathway.

The Association between Smoking and Epiretinal Membrane.

We conducted a meta-analysis of analytic and observational studies to evaluate the association between smoking and epiretinal membrane (ERM). The pertinent studies were identified via a literature search using three databases (MEDLINE, Cochrane Library, Embase) and the reference lists of retrieved studies. Cohort, case-control and cross-sectional studies meeting the predefined criteria were included. We extracted the odds ratios (OR) and 95% confidence intervals (CI) from each study. Overall risk estimates were pooled using random-effects models. Subgroup analyses based on several stratified factors were also performed. Two cohort studies and six cross-sectional studies involving 46,837 subjects were included. The pooled effect of all eight studies showed an unexpected significant decreased association between smoking and the occurrence of ERM (OR, 0.72; 95% CI 0.61-0.84; p = 0.29, I2 = 17.9%). Subgroup analyses supported this finding, except for the age-unadjusted group (OR, 0.87; 95% CI 0.63-1.22), the ERM classification group (cellophane macular reflex (CMR) OR, 0.93; 95% CI 0.68-1.28; preretinal macular fibrosis (PMF) OR, 0.74; 95% CI 0.41-1.32), the Asian group (OR, 0.75; 95% CI 0.52-1.09) and the past smoker group (OR, 1.02; 95% CI 0.85-1.22). The pooled effects from the current literature suggested a declining association between smoking and ERM, which requires further studies to confirm.